4-Substituted 1-chloro-2-nitrobenzenes: Structure-activity relationships and extension of the substrate model of rat glutathione S-transferase 4-4

Article abstract of DOI:10.1021/tx960137w

In the present study, eleven 4-substituted 1-chloro-2-nitrobenzenes were tested for their GSH conjugation capacity when catalyzed by base or rat glutathione S-transferase (GST) 4-4. Kinetic parameters (k(s) and K(m), k(cat), and k(cat)/K(m)) were determined and subsequently used for the description of structure-activity relationships (SAR's). For this purpose, eight physicochemical parameters (electronic, steric, and lipophilic) of the substituents and five computer-calculated parameters of the substrates (charge distributions and several energy values) were used in regression analyses with the kinetic parameters. The obtained SAR's are compared with corresponding SAR's for the GSH conjugation of 2-substituted 1-chloro-4- nitrobenzenes, previously determined [Van der Aar et al. (1996) Chem. Res. Toxicol. 9, 527-534]. The kinetic parameters of the 4-substituted 1-chloro- 2-nitrobenzenes correlated well with the Hammett σ(p)^- constant; the Hammett σ(p) constant corrected for 'through resonance', while the corresponding kinetic parameters of the 2-substituted 1-chloro-4- nitrobenzenes did not. The base-and GST 4-4-catalyzed GSH conjugation reactions of 2-substituted 1-chloro-4-nitrobenzenes depend to a different extent on the electronic properties of the ortho substituents, suggesting the involvement of different rate-limiting transition states. The base- and GST 4-4-catalyzed conjugation of 4-substituted 1-chloro-2-nitrobenzenes, however, showed a similar dependence on the electronic properties of the para substituents, indicating that these substrates are conjugated to GSH via a similar transition state. Multiple regression analyses revealed that, besides electronic interactions, also steric and lipophilic restrictions appeared to play an important role in the GST 4-4-catalyzed GSH conjugation of 4- substituted 1-chloro-2-nitrobenzenes. Finally, the 4-substituted 1-chloro-2- nitrobenzenes were also used to extend the previously described substrate model for GST 4-4 [De Groot et al. (1995) Chem. Res. Toxicol. 8, 649-658], by which a specific steric restriction of substrates for GST 4-4 became clear.

Full text of DOI:10.1021/tx960137w

Chem. Res. Toxicol. 1997, 10, 439-449
439
4-Su bstitu ted 1-Ch lor o-2-n itr oben zen es:
Str u ctu r e-Activity Rela tion sh ip s a n d Exten sion of th e
Su bstr a te Mod el of Ra t Glu ta th ion e S-Tr a n sfer a se 4-4
Ellen M. van der Aar,† Marcel J . de Groot,†,‡ Tialda Bouwman,†
Greetje J . Bijloo,†,‡ J an N. M. Commandeur,† and Nico P. E. Vermeulen*,†
Leiden/ Amsterdam Center for Drug Research, Department of Pharmacochemistry,
Divisions of Molecular Toxicology and Medicinal Chemistry, Vrije Universiteit, De Boelelaan 1083,
1081 HV Amsterdam, The Netherlands
Received August 1, 1996X
In the present study, eleven 4-substituted 1-chloro-2-nitrobenzenes were tested for their GSH
conjugation capacity when catalyzed by base or rat glutathione S-transferase (GST) 4-4. Kinetic
parameters (ks and Km, kcat, and kcat/Km) were determined and subsequently used for the
description of structure-activity relationships (SAR’s). For this purpose, eight physicochemical
parameters (electronic, steric, and lipophilic) of the substituents and five computer-calculated
parameters of the substrates (charge distributions and several energy values) were used in
regression analyses with the kinetic parameters. The obtained SAR’s are compared with
corresponding SAR’s for the GSH conjugation of 2-substituted 1-chloro-4-nitrobenzenes,
previously determined [Van der Aar et al. (1996) Chem. Res. Toxicol. 9, 527-534]. The kinetic
parameters of the 4-substituted 1-chloro-2-nitrobenzenes correlated well with the Hammett
-
σp constant; the Hammett σp constant corrected for “through resonance”, while the corre-
sponding kinetic parameters of the 2-substituted 1-chloro-4-nitrobenzenes did not. The base-
and GST 4-4-catalyzed GSH conjugation reactions of 2-substituted 1-chloro-4-nitrobenzenes
depend to a different extent on the electronic properties of the ortho substituents, suggesting
the involvement of different rate-limiting transition states. The base- and GST 4-4-catalyzed
conjugation of 4-substituted 1-chloro-2-nitrobenzenes, however, showed a similar dependence
on the electronic properties of the para substituents, indicating that these substrates are
conjugated to GSH via a similar transition state. Multiple regression analyses revealed that,
besides electronic interactions, also steric and lipophilic restrictions appeared to play an
important role in the GST 4-4-catalyzed GSH conjugation of 4-substituted 1-chloro-2-
nitrobenzenes. Finally, the 4-substituted 1-chloro-2-nitrobenzenes were also used to extend
the previously described substrate model for GST 4-4 [De Groot et al. (1995) Chem. Res. Toxicol.
8, 649-658], by which a specific steric restriction of substrates for GST 4-4 became clear.
gene class, but intergene class heterodimers are not
known. Each subunit contains a binding site for GSH
(G-site) and a separate binding site for electrophilic
substrates (H-site), and there have been considerable
efforts to characterize the molecular structure of these
binding sites. These approaches include site-directed
mutagenesis (6, 7) and construction of chimeric protein
subunits (8) as well as affinity labeling (9, 10), NMR
spectroscopy (11), and high-resolution X-ray studies on
crystals of GST isoenzymes (12-14).
In tr od u ction
The glutathione S-transferases (GSTs; EC 2.5.1.18)1
constitute a family of enzymes that catalyze the addition
of the tripeptide GSH to lipophilic substrates bearing
electrophilic functional groups (1, 2). The cytosolic
enzymes are dimeric proteins with subunits (22-27 kDa)
that appear to be organized into at least five principal
gene classes designated R, µ, π (3), θ (4), and σ (5). Both
homo- and heterodimeric enzymes are found within a
Recently, molecular modeling techniques were used to
derive a substrate model for rat GST 4-4, an isoenzyme
of which a crystal structure has not been described yet
(15). Information on regio- and stereoselective product
formation of 20 substrates, covering structurally different
classes of chemicals [aromatic diol epoxides, pyrene oxide,
(aza)phenanthrene oxides, and aromatic chlorides] was
used to construct a substrate model containing three
interaction sites responsible for Lewis acid-Lewis base
interactions (IS1, IS2, and IS3), as well as a region
responsible for aromatic interactions (IS4) (Figure 1).
These interaction sites in the substrates corresponded
with hypothetical protein interaction sites (pIS1, pIS2,
pIS3, and pIS4) in the active site of GST 4-4 (15).
* To whom correspondence should be addressed (tel, +31 (0)20 444
7590; fax, +31 (0)20 444 7610; e-mail, vermeule@chem.vu.nl).
† Division of Molecular Toxicology.
‡ Division of Medicinal Chemistry.
X Abstract published in Advance ACS Abstracts, April 1, 1997.
1 Abbreviations: BPDE, benzo[a]pyrene 7,8-diol 9,10-epoxide; CDNB,
1-chloro-2,4-dinitrobenzene; CSD, Cambridge Structural Database;
DMA, distributed multipole analysis; DMSO, dimethyl sulfoxide; Ehomo
,
energy of the highest occupied molecular orbital; Elumo, energy of the
lowest unoccupied molecular orbital; GAMESS-UK, Generalised Atomic
and Molecular Electronic Structure System, UK version; GS-, glu-
tathione anion; GS-conjugate, GSH conjugate; GSTs, glutathione
S-transferases; IS, interaction site in the substrate, which interacts
with a corresponding pIS in the enzyme; λmax, maximal absorbance
wavelength; MeS-, methanethiolate anion or model nucleophile for the
thiolate anion of GSH; pIS, protein interaction site, responsible for
interaction with a corresponding IS in the substrate; RHF, restricted
Hartree-Fock; STO, Slater-type orbital; SAR’s, structure-activity
relationships; SNAr, nucleophilic aromatic substitution; SV, split
valence.
Structure-activity relationships (SAR’s) have also
been used to characterize the mechanism of action and
S0893-228x(96)00137-3 CCC: $14.00 © 1997 American Chemical Society

440 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Van der Aar et al.
nitrobenzenes, by which the steric restrictions of sub-
strates for GST 4-4 became more clear.
Exp er im en ta l P r oced u r es
Ma ter ia ls a n d Ch em ica ls. GSH was obtained from Boe-
hringer (Mannheim, Germany), and 4-chloro-3-nitrobenzalde-
hyde (2b; 97% pure), 4-chloro-3-nitrobenzophenone (3b; 99%
pure), 1-chloro-2-nitro-4-trifluoromethylbenzene (8b; 98% pure),
and 1,4-dichloro-2-nitrobenzene (9b; 99% pure) were from
Aldrich-Chemie (Steinheim, Germany). 1-Chloro-2-nitroben-
zene and 4-chloro-3-nitrobenzoic acid were purchased from
Fluka Chemie AG (Buchs, Switzerland). Chemical reagents and
buffers were of the highest quality commercially available. Rat
liver GST 4-4 was purified as described before (20).
F igu r e 1. Substrate model for the active site of GST 4-4 with
benzo[a]pyrene 7(R),8(S)-diol 9(S),10(R)-epoxide (BPDE) as the
template (taken from 15). Epoxide oxygen atom (IS1), C8-
hydroxyl group (IS2), and C7-hydroxyl group (IS3), interacting
with protein interaction sites (pIS1, pIS2 and pIS3, respectively),
and aromatic interaction region (IS4) interacting with aromatic
amino acids in the protein (pIS4).
the active sites of GSTs. Chemical variation of the
structure of the cofactor GSH gave more insight in
important interactions between GSH and amino acids in
the G-site of GST 3-3 (16). The significance of hydro-
phobic and steric interactions in the active site of GSTs
was explored by the determination of substrate specifici-
ties of 15 cytosolic GSTs from rat, human, and mouse by
use of a homologous series of ten 4-hydroxyalkenals (17).
Recently, Rietjens et al. described parallel SAR’s for the
overall rate of GSH conjugation of a series of five
fluoronitrobenzenes when catalyzed by base and human
and rat cytosolic GSTs (18, 19). For all catalysts, the
interaction between the thiolate anion of GSH and the
fluoronitrobenzene, leading to the Meisenheimer inter-
mediates, is the rate-limiting step in overall conversion
of this series of small substrates. Comparison of enzyme
kinetic parameters of the GSH conjugation of eleven
2-substituted 1-chloro-4-nitrobenzenes, when catalyzed
by rat GST isoenzymes of three different classes, showed
major differences in the enzyme kinetic parameters and
among others, in the GST isoenzyme selectivity (20).
Recently, we also described SAR’s for the GSH conjuga-
tion of a series of ten 2-substituted 1-chloro-4-nitroben-
zenes catalyzed by base and by rat GST 4-4 (21). Steric,
lipophilic, and electronic parameters were correlated with
the kinetic parameters of the GSH conjugation reaction
of these substrates. Moreover, charge distributions and
several other molecular parameters were calculated for
the substrates and the corresponding Meisenheimer
intermediates with MeS- as a model nucleophile for the
thiolate anion of GSH and used in regression analyses.
The comparison of SAR’s for the base- and GST 4-4-
catalyzed GSH conjugation reaction led, i.a., to the
proposal that the rate-limiting transition states for the
base-catalyzed and the GST 4-4-catalyzed GSH conjuga-
tion reaction are different (21).
Syn th eses. Methyl 4-chloro-3-nitrobenzoate (4b) and n-butyl
4-chloro-3-nitrobenzoate (5b) were synthesized by esterification
of 4-chloro-3-nitrobenzoic acid with the corresponding alcohols.
In short, a mixture of 4 g (20 mmol) of 4-chloro-3-nitrobenzoic
acid, 1.8 mL of sulfuric acid (95-98%), and 120 mL of methanol
or n-butanol for 4b and 5b, respectively, was refluxed while
stirring for 18 h. After the solution had cooled to room
temperature, 75 mL of water was added. For purification of
substrate 4b, the methanol was evaporated and the residual
water phase was neutralized with sodium bicarbonate, in which
4b is insoluble. The crystals were filtered, washed with cold
water, and dried. Recrystallization from methanol gave white
crystals of 4b in a yield of 96%, mp 80 °C. 1H-NMR (2H-
DMSO): δ 3.9 (s, CH3, 3H), 7.9 (d, H5, J 56 ) 8.0 Hz, 1H), 8.2
(dd, H6, J 56 ) 8.0 Hz, J 26 ) 1.33 Hz, 1H), 8.5 (d, H2, J 26 ) 1.33
Hz, 1H). For purification of 5b the n-butanol was slowly
evaporated. The residual water phase, which contained oily
drops, was extracted with ethyl acetate, and subsequently the
ethyl acetate fractions were washed with a 5% (w/v) sodium
bicarbonate solution to remove unreacted 4-chloro-3-nitrobenzoic
acid. The ethyl acetate was dried over sodium sulfate and
evaporated in vacuo. The yellow oil (yield 45%) appeared to be
pure as judged from NMR and HPLC [assay used for determi-
nation of extinction coefficient (20)]. 1H-NMR (2H-DMSO): δ
0.9 (t, CH3, J ) 6.67 Hz, 3H), 1.42 (m, CH2, 2H), 1.70 (m, CH2,
2H), 4.3 (t, CH2, J ) 6.67 Hz, 2H), 7.9 (d, H5, J 56 ) 8.0 Hz, 1H),
8.2 (dd, H6, J 56 ) 8.0 Hz, J 26 ) 1.33 Hz, 1H), 8.5 (d, H2, J 26
)
1.33 Hz, 1H).
tert-Butyl 4-chloro-3-nitrobenzoate (6b) was synthesized by
adding 4 g (20 mmol) of 4-chloro-3-nitrobenzoic acid to 25 mL
of dichloromethane and cooling this mixture to -20 °C in a flask
able to resist 2 atm of pressure. With the aid of a CO2/acetone
mixture 20 mL of isobutylene was condensed. The liquid
isobutylene was subsequently added to the dichloromethane
mixture via a calcium chloride tube. After addition of 0.2 mL
of sulfuric acid (95-98%), the flask was closed and shaken for
2 days at room temperature. After the reaction mixture had
been cooled, it was poured into 300 mL of a saturated sodium
bicarbonate solution. This mixture was extracted three times
with dichloromethane. The dichloromethane fractions were
washed with a 5% (w/v) sodium bicarbonate solution, water, and
a saturated sodium chloride solution, respectively. To remove
polymers, the dichloromethane phase was eluted over a small
silica column, dried over sodium sulfate, and finally the dichlo-
romethane was evaporated. The residue was purified over a
silica column, eluted with a mixture of petroleum ether (40/60)
and ether (10:1). The yield of the white crystalline product was
3 g (58%), mp 68 °C. 1H-NMR (2H-DMSO): δ 1.6 (s, CH3, 9H),
In the present study the kinetic parameters of the
base-catalyzed (ks) and GST 4-4-catalyzed (Km, kcat, and
k
cat/Km) GSH conjugation of eleven 4-substituted 1-chloro-
2-nitrobenzenes (Table 1) are determined, compared, and
described in terms of SAR’s using eight classical physi-
cochemical parameters and five computer-calculated
molecular parameters of the substrates. The obtained
SAR’s were compared with SAR’s for the base- and the
GST 4-4-catalyzed GSH conjugation of 2-substituted
1-chloro-4-nitrobenzenes (Table 1), determined in a pre-
vious study (21). These SAR equations and their differ-
ences gave more detailed information about the active
site of GST 4-4 and which interactions (electronic, steric,
lipophilic) are important for the GSH conjugation of these
substrates by GST 4-4. Also the previously reported
substrate model for rat GST 4-4 (15) is refined by
including the present series of 4-substituted 1-chloro-2-
7.6 (d, H5, J 56 ) 10.0 Hz, 1H), 8.1 (dd, H6, J 56 ) 10.0 Hz, J 26
3.34 Hz, 1H), 8.4 (d, H2, J 26 ) 3.34 Hz, 1H).
)
Synthesis of 1-bromo-4-chloro-3-nitrobenzene (10b) was ac-
complished according to Hammond and Modic (22). To a
mixture of 1 g (6.35 mmol) of 1-chloro-2-nitrobenzene, 1 g of
Ag2SO4, and 10 mL of sulfuric acid (95-98%) was slowly added
0.4 mL of bromine. This mixture was heated on a water bath
of 80 °C for 5 h under constant stirring. After cooling, water
was added to the reaction mixture and subsequently it was
extracted with diethyl ether. The ether extract was dried over

4-Substituted 1-Chloro-2-nitrobenzenes and Rat GST 4-4
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 441
Ta ble 1. R Gr ou p s in 2-Su bstitu ted 1-Ch lor o-4-n itr oben zen es (a -Com p ou n d s) a n d 4-Su bstitu ted
1-Ch lor o-2-n itr oben zen es (b-Com p ou n d s) a n d Th eir Refer en ce Cod esa
λmax (nm)
GS-conjugates b
ꢀ (mM-1 cm-1
GS-conjugate b
)
R
substrates b
NO2
CHO
1a
2a
3a
4a
5a
6a
7a
8a
9a
10a
1b
2b
3b
4b
5b
6b
250
245
260
232
233
232
340b
304c
306
370
370
289
9.6b
10.0c
COC6H5
CO2CH3
CO2(CH2)3CH3
CO2C(CH3)3
CN
CF3
Cl
Br
14.1 ( 0.4
3.4 ( 0.02
2.9 ( 0.1
16.4 ( 1.2
8b
9b
10b
11b
12b
253
221
223
271
230
365c
383d
380
294
282
2.2c
0.7d
1.70 ( 0.01
18.8 ( 4.5
e
CO2C6H5
CO2C6H4(tBu)
a
The maximal absorbance wavelength (λmax) of the 4-substituted 1-chloro-2-nitrobenzenes and their corresponding GS-conjugates are
shown, as well as the extinction coefficients (ꢀ) of the GS-conjugates, which are determined at the λmax of the GS-conjugates [determined
b
according to Van der Aar et al. (20)] and are the mean ( SD of at least three independent experiments. Taken from Habig and J akoby
(43). c Taken from Chen et al. (41). Taken from Keen et al. (40). e Not determined.
d
magnesium sulfate, and the diethyl ether was evaporated.
Recrystallization of the solid residue from ethanol gave light
yellow crystals in
determined for the GSH conjugation of 4-substituted 1-chloro-
2-nitrobenzenes (b-compounds, Table 1) when catalyzed by base
and GST 4-4. In short, the formation of GS-conjugates was
followed spectrophotometrically with time at the maximal
absorbance wavelength (λmax) of the GS-conjugates (Table 1).
For all examined substrates no appreciable absorbance of
reactants was observed at the wavelengths that were used to
monitor the formation of GS-conjugates. The extinction coef-
ficients (ꢀ) of the GS-conjugates were determined at the wave-
lengths that were used to follow the reactions according to Van
der Aar et al. (20) and are shown in Table 1. The assay medium
for determination of GST-catalyzed conjugation rates contained
0.1 M potassium phosphate buffer (pH 6.5, 0.1 mM EDTA), GST
4-4 (different concentrations depending on substrate tested), 1
mM GSH, and various concentrations of substrate. The incuba-
tions were performed at 37 °C. The base-catalyzed GSH
conjugation reaction was performed in 0.1 M potassium phos-
phate buffer (pH 9.2, 0.1 mM EDTA), at 50 °C with a GSH
concentration of 10 mM. Under these experimental conditions
no GSH was oxidized, as determined with 2,2′-dinitro-5,5′-
dithiodibenzoic acid.
Com p u ta tion a l Meth od s a n d Molecu la r Mod elin g. The
minimal energy conformers of the compounds were calculated,
and it was assumed that these minimal energy conformers were
those that would bind to the enzyme. This might, however, not
be the actual situation. The initial conformations of the
4-substituted 1-chloro-2-nitrobenzenes (Table 1, 1b-12b) were
generated from the geometry-optimized conformation (15) of
1-chloro-2,4-dinitrobenzene (CDNB, 1a ,b) using the molecular
modeling package ChemX (23). Macromodel (24) was used for
molecular mechanics conformational analyses on the generated
compounds. If necessary, a large number of conformations was
generated by changing all rotatable bonds with increments of
30° (compounds 5b, 6b, 11b, and 12b). Generally, the confor-
mations were energy minimized with the Amber force field (25,
26), using Batchmin (27). For the ketone- and ester-substituted
substrates several low-energy conformations were generated
with different orientations of the carbonyl oxygen atom (see
Figure 2).
a
yield of 29%, mp 69 °C. 1H-NMR
(2H-chloroform): δ 7.4 (d, H5, J 56 ) 8.0 Hz, 1H), 7.6 (dd, H6, J 56
) 8.0 Hz, J 26 ) 0.67 Hz, 1H), 8.0 (d, H2, J 26 ) 0.67 Hz, 1H).
Phenyl 4-chloro-3-nitrobenzoate (11b) was synthesized by
converting 4-chloro-3-nitrobenzoic acid to the corresponding
benzoyl chloride, followed by a reaction with phenol. 4-Chloro-
3-nitrobenzoic acid (1 g, 5 mmol) was mixed with 1 g (5 mmol)
of phosphorus pentachloride and refluxed for 2 h at 80 °C under
a nitrogen atmosphere. After cooling the reaction mixture,
phosphorus oxychloride was removed by distillation under
reduced pressure (25 °C/20 mmHg). The residual 4-chloro-3-
nitrobenzoyl chloride was used without further purification.
Phenol, 0.5 g (5 mmol), was dissolved in 5 mL of dry pyridine
and added to the benzoyl chloride derivative. This reaction
mixture was refluxed at 90 °C for 40 min. After the mixture
had been cooled, it was poured into 20 mL of 2 M hydrochloric
acid. The aqueous phase was decanted from the precipitated
oil and stirred vigorously with 10 mL of a saturated sodium
carbonate solution. The resulting solid material was filtered
and washed with cold water. Recrystallization from ethanol
gave yellow-orange crystals in a yield of 4%, mp 70 °C. 1H-
NMR (2H-DMSO): δ 7.4 (m, C6H5, 3H), 7.5 (m, C6H5, 2H), 8.1
(d, H5, J 56 ) 6.67 Hz, 1H), 8.4 (dd, H6, J 56 ) 6.67 Hz, J 26 ) 3.34
Hz, 1H), 8.7 (d, H2, J 26 ) 3.34 Hz, 1H).
3′-tert-Butylphenyl 4-chloro-3-nitrobenzoate (12b) was syn-
thesized by reacting 4-chloro-3-nitrobenzoyl chloride with 3-tert-
butylphenol (as described for the synthesis of 4-chloro-3-
nitrophenylbenzoate). After cooling the reaction mixture, 20 mL
2 M hydrochloric acid was added after which the pH was
adjusted to 6-7 with a saturated sodium carbonate solution.
This solution was extracted with diethyl ether and the collected
diethyl ether fractions were washed five times with 2 M sodium
hydroxide to remove unreacted 3-tert-butylphenol. The diethyl
ether fraction was dried over magnesium sulfate and evaporated
in vacuo. 3′-tert-Butylphenyl 4-chloro-3-nitrobenzoate was ob-
tained as a light yellow oil in a yield of 5%. 1H-NMR (2H-
DMSO): δ 1.4 (s, CH3, 9H), 7.1-7.5 (m, C6H4, 4H), 8.1 (d, H5,
J 56 ) 8.0 Hz, 1H), 8.4 (dd, H6, J 56 ) 8.0 Hz, J 26 ) 2.0 Hz, 1H),
8.8 (d, H2, J 26 ) 2.0 Hz, 1H).
The quantum chemical program package GAMESS-UK (28,
29) was used for the ab initio calculations described. The
geometries of the lowest energy conformations for all substrates
obtained with Macromodel, were ab initio optimized at the RHF
(restricted Hartree-Fock) level using the STO-3G (Slater type
orbital composed of 3 Gaussians) (30) minimal basis set. On
the resulting STO-3G geometries, a single-point energy calcula-
tion and a DMA (distributed multipole analysis) calculation (31)
Kin etic P a r a m eter s. The determination of the enzyme
kinetic parameters Km, kcat, and kcat/Km for the GST 4-4-
catalyzed GSH conjugation of 2-substituted 1-chloro-4-nitroben-
zenes (a -compounds, Table 1) and the rate constants (ks) of the
corresponding base-catalyzed reactions has been described in
previous studies (20, 21). The same kinetic parameters were

442 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Van der Aar et al.
Ta ble 2. Kin etic P a r a m eter s of th e GST 4-4-Ca ta lyzed
(Km , kca t, kca t/Km ) a n d th e Ba se-Ca ta lyzed (ks) GSH
Con ju ga tion of 4-Su bstitu ted 1-Ch lor o-2-n itr oben zen es
(b-Com p ou n d s)
ks
(×10-4
kcat/Km
Km
(µM)
kcat
(×10-2 µM-1
substrate µM-1 min-1
)
(min-1
)
min-1
)
1b
2b
3b
4b
5b
6b
8b
9b
420 ( 20
23 ( 1
156 ( 36 137 ( 14
93 ( 19
35 ( 8
21 ( 4
2.7 ( 0.8
3.4 ( 0.3
a
0.7 ( 0.1
b
b
73 ( 6
41 ( 5
31 ( 9
F igu r e 2. Low-energy conformations of 1-chloro-2-nitroben-
32 ( 1
6.5 ( 1.0
308 ( 57 8.5 ( 0.7
zenes with ketone or ester substituents on the 4-position.
7.0 ( 0.7
10.7 ( 1.2 50 ( 5
1.70 ( 0.01
a
were performed using the RHF method in an SV (split valence)
6-31G (32, 33) basis set. Charge distributions and energy values
(Ehomo and Elumo) of the lowest total energy conformation of the
ketone- and ester-substituted substrates were used in the SAR
analyses (energy difference between various conformations <4
kJ /mol).
2.8 ( 0.4
a
5.3 ( 0.3
748 ( 180 7.3 ( 2.0
b
b
b
b
b
b
10b
11b
12b
40.2 ( 3.8 46 ( 13
c
6.8 ( 1.5
d
15 ( 2
d
d
a
The molecular modeling program ChemX (23) was used for
the rigid fitting procedure in which compounds are fitted
allowing only global rotations and translations. All 4-substi-
tuted 1-chloro-2-nitrobenzenes (1b-12b) were fitted onto the
diequatorial conformation of benzo[a]pyrene 7(R),8(S)-diol 9(S),-
10(R)-epoxide (BPDE) with the rigid fitting procedure (23). The
sites of GSH attack in the different substrates were matched
onto each other. Furthermore, one of the oxygen atoms of the
nitro group at the 2-position of the substrates was fitted onto
the oxygen atom of the 8-hydroxyl group of BPDE. If the
distance between fitted atoms was less than 0.25 Å, the fitted
conformation was accepted. The energy of the fitted conforma-
tion was calculated with ab initio methods [RHF/SV 6-31G (32,
33)]. If the energy difference between the minimal energy
conformation and the fitted conformation was less than 21
kJ /mol, the fitted conformation was accepted and a DMA (31)
was carried out.
Str u ctu r e-Activity Rela tion sh ip s. Single- and multiple-
regression analyses were performed for determination of cor-
relations between kinetic parameters of the GST 4-4-catalyzed
(Km, kcat, and kcat/Km) and the base-catalyzed (ks) GSH conjuga-
tion reaction of 4-substituted 1-chloro-2-nitrobenzenes (1b-12b)
with eight physicochemical parameters and five computer-
calculated parameters. These correlations were compared with
those for the 2-substituted 1-chloro-4-nitrobenzenes (1a -10a )
previously determined (21).
To probe the steric effects of the substituents, the multidi-
mensional Sterimol parameters (L, B1, and B5) were used (34).
The hydrophobic fragment constant f of Rekker (35) was used
to probe lipophilicity effects of the substituents. The Hammett
σp constants were used as electronic parameters (taken from
ref 36). Despite the steric influences in the ortho substituents,
the σ values for para substituents (σp) approximate those for
the ortho substituents in the case of 2-substituted 1-chloro-4-
nitrobenzenes (21). When substituents are directly conjugated
with the reaction center, “through resonance” may occur. Such
interactions may either facilitate or hinder attainment of the
transition state. Therefore, an alternative Hammett constant,
σ-, has been defined in order to correct for “through resonance”
(37). Furthermore, two additional electronic parameters were
defined (38) to separate the inductive component of substituents
(field effect, F ) from the resonance component (R ) (taken from
ref 36). The physicochemical parameters actually used in the
SAR analyses are shown in Table 3.
To obtain the computer-calculated molecular parameters of
the substrates (1b-12b), the lowest energy conformations of
the substrates were subjected to energy and charge distribution
calculations, leading to Ehomo and Elumo, to charges on the
attacked C-atom (C1), and to charges on the ortho and para
substituents, respectively. For the ten 2-substituted 1-chloro-
4-nitrobenzenes (1a -10a ) these parameters were calculated and
tabulated in a previous study (21), while the molecular param-
eters of the 4-substituted 1-chloro-2-nitrobenzenes were calcu-
lated in this study and are shown in Table 3.
No detectable GSH conjugation up to 300 µM of substrate 6b
(kcat < 0.06 min-1 using 250 µg/mL of GST 4-4). No detectable
b
GSH conjugation up to 1000 µM of substrates 9b and 10b (kcat
<
1.5 and 0.6 min-1, respectively using 250 µg/mL of GST 4-4). c GSH
conjugation observed but not quantified (no extinction coefficient
d
determined). No detectable GSH conjugation up to 300 µM of
substrate 12b (no kcat limit calculated due to the unavailability of
the extinction coefficient).
intercorrelation between the two independent parameters was
checked (r < 0.5).
Resu lts a n d Discu ssion
In the present study eleven 4-substituted 1-chloro-2-
nitrobenzenes (compounds 1b-12b, Table 1) were ana-
lyzed for their capacity to conjugate to GSH when
catalyzed by GST 4-4 or base. When possible, the
corresponding kinetic parameters (Km, kcat, kcat/Km, and
ks, respectively) of these compounds were determined and
compared (Table 2). SAR analyses were performed by
correlating the kinetic parameters for the GSH conjuga-
tion of 4-substituted 1-chloro-2-nitrobenzenes with eight
physicochemical parameters (σp, σp-, F , R , L, B1, B5, and
f) and with five computer-calculated molecular param-
eters of the substrates (Elumo, Ehomo, and the charges on
C1 and on the ortho and para substituents) (Table 3).
Correlations were only considered to be significant when
t > |2|. These SAR’s are compared with the previously
obtained SAR’s for the base- and GST 4-4-catalyzed GSH
conjugation of ten 2-substituted 1-chloro-4-nitrobenzenes
(compounds 1a -10a , Table 1) (21).
Furthermore, eight 4-substituted 1-chloro-2-nitroben-
zenes and three 2-substituted 1-chloro-4-nitrobenzenes
are used to extend the previously described substrate
model for rat GST 4-4, which was obtained by molecular
modeling techniques and which was among others based
on some 2-substituted 1-chloro-4-nitrobenzenes (1a -5a ,
9a , and 10a ) (Figure 1) (15).
Ap p a r en t Mich a elis Con sta n t (Km ). The experi-
mentally determined apparent Km values of 4-substituted
1-chloro-2-nitrobenzenes for GST 4-4 are shown in Table
2. The lowest Km values, suggestive of a high apparent
affinity, were found for substrates with a phenyl sub-
stituent (3b and 11b). The highest Km value by far was
found for the CF3-substituted substrate 8b. For the p-Cl-
and p-Br-substituted compounds (9b and 10b,) no detect-
able GST 4-4-catalyzed GSH conjugation was observed
up to 1 mM (Table 2, kcat < 1.5 and 0.6 min-1, respec-
tively, when using 250 µg/mL of GST 4-4). On the basis
of the the higher GSH conjugation rates of most 4-sub-
stituted 1-chloro-2-nitrobenzenes when compared to the
Only correlations with a Student’s t test’s t value of >|2| were
considered to be significant. In case of multiple regression the

4-Substituted 1-Chloro-2-nitrobenzenes and Rat GST 4-4
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 443
Ta ble 3. Com p u ter -Ca lcu la ted Molecu la r P a r a m eter sa of 4-Su bstitu ted 1-Ch lor o-2-n itr oben zen es Exa m in ed in Th is
Stu d y a n d th e P h ysicoch em ica l P a r a m eter s of th e Su bstitu en ts Actu a lly Used in Th is Stu d y
chargeb
c
c
attacked
C atom
o-nitro
substituent
para
substituent
Elumo
Ehomo
-
substrates
σp
L
B5
f
(kJ /mol)
(kJ /mol)
1b
2b
3b
4b
5b
6b
8b
9b
10b
11b
1.27
1.02
0.87
0.76
0.76
d
0.71
0.21
0.23
0.76
3.44
3.53
4.57e
4.73
8.00
d
3.30
3.52
3.82
8.13
2.44
2.36
5.98e
3.36
5.85
d
2.61
1.80
1.95
3.50
-0.039
-0.333
0.926
0.181
1.738
1.738
1.223
0.933
1.134
1.590
0.265
0.179
0.173
0.184
0.183
0.182
0.185
0.168
0.148
0.185
-0.148
-0.171
-0.179
-0.172
-0.172
-0.173
-0.148
-0.177
-0.180
-0.169
-0.169
0.142
0.188
0.080
0.086
0.097
0.093
-0.147
-0.008
0.079
44.3
17.5
36.5
37.0
39.4
42.1
-8.5
26.0
37.1
32.3
-1094.4
-1027.5
-946.3
-1021.6
-1017.6
-1013.8
-1060.4
-996.3
-959.7
-913.8
a
b
The computer-calculated parameters were taken from the conformation with the lowest total energy. Charges were obtained from
DMA calculations. c Energy values were obtained from SV 6-31G calculations. This physicochemical parameter was not available for a
d
tert-butyl ester substituent. e Phenyl ring is twisted out of the aromatic plane (90°).
2-substituted 1-chloro-4-nitrobenzenes (discussed later),
it was expected that 9b and 10b would also be conjugated
to GSH. Unfortunately, however, the extinction coef-
ficients of the corresponding GS-conjugates of 9b and 10b
are low (0.7 and 1.7 mM-1 cm-1, Table 1), thereby
prohibiting a sensitive and accurate measurement. In-
stead, in an incubation mixture of GST 4-4 and GSH, the
decrease in substrates 9b and 10b was checked with time
by HPLC with UV detection at λmax of the substrates.
After 30 min, a measurable decrease in concentrations
of substrates was observed ((8%) when compared to the
control incubations in which no GSH was present,
indicating that some GSH conjugation may have taken
place (data not shown). Detectable GSH conjugation was
not observed for substrates 6b and 12b up to 300 µM.
This could also be a result of the limited solubility of
these substrates or to steric restrictions in this area of
the active site of GST 4-4. Electronic restrictions prob-
ably play no role because measurable GSH conjugation
of 6b and 12b was observed when catalyzed by base
(Table 2).
Single- and multiple-regression analyses between the
Km values of GST 4-4 for the GSH conjugation of
4-substituted 1-chloro-2-nitrobenzenes (compounds 1b-
12b) (Table 2) with all eight physicochemical and five
computer-calculated parameters discussed before, showed
no statistically significant correlations (t < |2|) or cor-
relations with a low correlation coefficient (r < 0.7).
Previously, significant correlations were neither found
for the Km values of 2-substituted 1-chloro-4-nitroben-
zenes (1a -10a ) with the above mentioned parameters
(21). Apparently, the affinity of the examined substrates
(1a -10a and 1b-12b) toward GST 4-4 (i.e., formation
of enzyme-substrate complexes) is not solely dependent
on any of these molecular parameters.
substituents, giving rise to additional steric and elec-
tronic effects.
A qualitative comparison of the apparent Km values of
GST 4-4 for the 4-substituted 1-chloro-2-nitrobenzenes
and the previously determined Km values for the 2-sub-
stituted series of 1-chloro-4-nitrobenzenes (20) shows that
there are significant differences. Substrates with bulky
substituents at the para position have a lower apparent
Km. Substrate 3b, for example, has an almost four times
lower Km than substrate 3a (41 and 154 µM, respectively).
From the calculated minimal energy conformation it
appeared that in both 3a and 3b, the phenyl ring of the
benzophenone substituent is twisted out of the aromatic
plane, due to intramolecular steric hindrance. The
difference in Km values of 3a and 3b could well be due to
this difference in orientation of the benzophenone phenyl
rings relative to the aromatic plane. The para substitu-
ent in 3b could also provide additional lipophilic and/or
electronic interactions with amino acids in the active site
of the GST 4-4 protein, which are absent in the case of a
NO2 substituent in 3a . The para-phenyl ester-substi-
tuted substrate (11b) also showed a low Km value (46
µM), again suggesting that there are additional lipophilic
and/or electronic interactions between the phenyl ring
and GST 4-4. Substrate 5b (n-butyl ester) has a Km
almost seven times lower than substrate 5a (50 and 341
µM, respectively). Steric restrictions in the active site
in accommodating the ortho-substituted n-butyl ester of
5a are probably not the reason for this, because the
corresponding methyl ester derivative (4a ) has an even
higher Km (554 µM). The high affinity of 5b for GST 4-4
is therefore more likely explained by lipophilic interac-
tions with the protein. The para-substituted methyl
ester derivative (4b) has a six times higher Km, and thus
a lower affinity than the corresponding para n-butyl ester
derivative (5b), probably because a methyl group provides
less lipophilicity than the flexible n-butyl chain in 5b
(Table 2).
Recently, Rietjens et al. found a linear correlation
between the Km values for the GSH conjugation of five
fluoronitrobenzenes catalyzed by rat liver cytosol and log
P
octanol (r ) -0.997) (18). In a successive study individual
Tu r n over Nu m ber kcat. Of the 4-substituted 1-chloro-
2-nitrobenzenes, the NO2- and CHO-substituted com-
pounds (1b and 2b) result in the highest and second
highest kcat values (Table 2). A comparison of the kcat
values of the 2-substituted 1-chloro-4-nitrobenzenes (1a -
10a ) (20) and the 4-substituted 1-chloro-2-nitrobenzenes
(1b-12b) (Figure 3) shows that substrates 2b-4b have
higher kcat values than their ortho-substituted analogues,
albeit to different extents. The n-butyl ester substrate
5a has a kcat seven times higher than 5b (12.1 ( 0.8 and
1.70 ( 0.01 min-1, respectively), and the o- and p-CF3-
GSTs of rat (1-1 and 3-3) and man (A1-1 and M1a-1a)
were subjected to similar regression analyses with four
fluoronitrobenzenes with varying numbers of F atoms
(19). For rat GST 1-1, a linear correlation was found
between the Km and log Poctanol (r ) -0.989), while for
rat GST 3-3 this correlation was less significant (r )
-0.874), indicating that correlations between Km and log
P depend on the individual GSTs. The fact that no such
correlations were found in the present study could well
be due to the inclusion of structurally more different

444 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Van der Aar et al.
F igu r e 3. Comparison of kcat values for the GST 4-4-catalyzed
GSH conjugation of 2-substituted 1-chloro-4-nitrobenzenes (a -
compounds, 9) and 4-substituted 1-chloro-2-nitrobenzenes (b-
compounds, 0).
F igu r e 4. Correlation between σp values of the 2-substituents
and the calculated charge on p-NO2 of 2-substituted 1-chloro-
4-nitrobenzenes (1a -10a ). The corresponding equation is as
follows: (charge p-NO2) ) 0.060((0.007)σp - 0.221((0.003), r
) 0.952, s ) 0.003, n ) 10.
substituted substrates 8a and 8b have almost identical
kcat values ((8 min-1).
tuted 1-chloro-4-nitrobenzenes (1a -10a ) was correlated
with the calculated charge on the p-NO2 substituents of
the substrates:
All substrates under investigation will be conjugated
to GSH via a nucleophilic aromatic substitution (SNAr)
mechanism. The rate-determining step in this type of
reactions is the formation of a Meisenheimer intermedi-
ate, in which both the nucleophile (GS-) and the leaving
group (Cl-) are bound to the substrate and in which a
net charge of -1 is distributed over the intermediate (39).
Electron-withdrawing substituents ortho and/or para
relative to the reaction center will stabilize this net
negative charge in the Meisenheimer intermediates.
Single-regression analyses between log kcat of GST 4-4
for the GSH conjugation of 2-substituted 1-chloro-4-
nitrobenzenes and 4-substituted 1-chloro-2-nitrobenzenes
and σp-, the Hammett electronic constant corrected for
“through resonance”, showed no significant correlation
for the 2-substituted analogues (t < 2), while the 4-sub-
stituted analogues did:
log kcat ) 41.8((7.2)(charge p-NO2) + 9.1((1.4) (2)
r ) 0.922
s ) 0.226
n ) 8
The CHO-substituted substrate (2a ) was excluded from
eq 2 because it appeared to be an outlier according to
statistical rules. Equation 2 implies that when the
negative charge on p-NO2 increases, the kcat increases.
The amount of negative charge on p-NO2 indirectly
reflects the electron-withdrawing capacity of the ortho
substituent (σo). This is illustrated in Figure 4, in which
the relation between σp of the 2-substituents and the
calculated negative charges on the p-NO2 substituent in
2-substituted 1-chloro-4-nitrobenzenes is shown. With
the corresponding equation (see legend to Figure 4) and
known σp values, it is possible to predict the charge on
the p-NO2 substituent and thus σo. When a regression
was applied between kcat and the charge on the o-NO2
substituent in 4-substituted 1-chloro-2-nitrobenzenes
(1b-12b), in contrast, no statistically significant cor-
relation was found (t ) 1.9). Apparently, the charge on
o-NO2 is not an appropriate parameter to describe the
electron-withdrawing capacity of the para substituents,
probably by the lack of “through resonance” on the o-NO2
substituent due to steric hindrance.
log kcat ) 2.77((0.67)σp- - 1.41((0.61)
r ) 0.901 s ) 0.319 n ) 6
(1)
The latter equation indicates that the higher the electron-
-
withdrawing capacity of the substituents becomes (σp
increases), the more efficient the Meisenheimer inter-
mediates are stabilized and the higher the kcat values
become. Within the present series of substituents, NO2
has the highest σp-, and thus the highest kcat value is
observed for substrate 1b. From an electronic point of
view the 4-substituted 1-chloro-2-nitrobenzenes should
have a higher GSH conjugation rate than the 2-substi-
tuted 1-chloro-4-nitrobenzenes. The para-substituted
analogues will have a more efficient stabilization of the
negative charge in the Meisenheimer complexes because
the ortho NO2 substituent is closest to the reaction center.
Exceptions to this general statement are possibly due to
extra steric and lipophilic interactions between the ortho
and/or para substituents and amino acids in the active
site. The fact that kcat for the 2-substituted 1-chloro-4-
Multiple-regression analyses (although with a limited
number of data points) with kcat of the 4-substituted
1-chloro-2-nitrobenzenes (1b-12b) and σp- and the Steri-
mol parameters L or B5 simultaneously revealed higher
correlation coefficients than with the individual param-
eters (r ) 0.90 for σp-, 0.74 for L, and 0.72 for B5):
log kcat
)
-0.16((0.05)L + 2.21((0.37)σp- - 0.17((0.47) (3)
r ) 0.982 s ) 0.161 n ) 6
-
nitrobenzenes (1a -10a ) did not correlate with σp indi-
log kcat
)
cates that “through resonance” on the ortho position is
not taking place, probably because of steric hindrance in
the Meisenheimer intermediates between the ortho sub-
stituent and the Cl atom at C1, while in the 4-substituted
analogues (1b-12b) this steric hindrance apparently
plays no siginificant role.
-0.16((0.05)B5 + 2.23((0.38)σp- - 0.31((0.45) (4)
r ) 0.981 s ) 0.165 n ) 6
These equations support our suggestion that, in addi-
tion to the electron-withdrawing capacity of the substit-
uents and the possibility for “through resonance”, also
A relatively high correlation coefficient was found when
the kcat of GST 4-4 for the GSH conjugation of 2-substi-

4-Substituted 1-Chloro-2-nitrobenzenes and Rat GST 4-4
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 445
steric factors play a role in the GST 4-4-catalyzed GSH
conjugation of 4-substituted 1-chloro-2-nitrobenzenes. In
the Sterimol system, L is defined as the length of the
substituent along the axis of the bond between the first
atom of the substituent and the parent molecule and B5
is the maximal width (34). Apparently, the L and B5
dimensions of the 4-substituents are limited and if these
dimensions are exceeded, no GSH conjugation is possible
anymore. This is probably the case for compound 12b,
which is conjugated to GSH when catalyzed by base but
not by GST 4-4. Compared to the GST 4-4 substrate 11b,
12b has an identical L value but a much larger B5 value,
thus preventing GSH conjugation.
Significant correlations were also found between log
kcat of GST 4-4 for the GSH conjugation of 4-substituted
1-chloro-2-nitrobenzenes (1b-12b) and the hydrophobic
fragment constant f of Rekker (35) (r ) 0.828) and the
calculated charge on the reaction center, C1 (r ) 0.769).
Multiple-regression analyses with both independent pa-
rameters, however, showed a higher correlation coef-
ficient:
F igu r e 5. Comparison of ks values for the base-catalyzed GSH
conjugation of 2-substituted 1-chloro-4-nitrobenzenes (b) and
4-substituted 1-chloro-2-nitrobenzenes (O) and of kcat/Km values
for the GST 4-4-catalyzed GSH conjugation of 2-substituted
1-chloro-4-nitrobenzenes (2) and 4-substituted 1-chloro-2-ni-
trobenzenes (4).
4-nitrobenzenes and 4-substituted 1-chloro-2-nitroben-
zenes (Figure 5) shows higher kcat/Km values for the latter
series of substrates, except for the n-butyl ester analogues
(5a and 5b), which have identical values, and for the CF3-
substituted substrates (8a and 8b), from which the ortho
analogue has the most efficient GSH conjugation. As
discussed before, in the 4-substituted 1-chloro-2-nitroben-
zenes (1b-12b), the NO2-substituent, with the highest
electron-withdrawing capacity, is positioned ortho rela-
tive to the GSH conjugation site (C1) and thereby is
closest to the reaction center, leading to an increased rate
of conjugation compared to the corresponding 2-substi-
tuted 1-chloro-4-nitrobenzenes (1a -10a ), in which the
NO2-substituent is located para relative to C1.
log kcat ) 10.09((3.13)(charge C1) - 0.51((0.14)f -
0.57((0.65) (5)
r ) 0.964
s ) 0.225
n ) 6
Although the regression analysis was performed with a
limited number of data points, there is a trend that the
lipophilicity of the para substituents is of importance in
the height of the value of kcat. The lipophilic properties
of the para substituents should not be too high, because
this seems to have a negative influence on kcat
. This
statement is confirmed by the observation that the para-
substituted n-butyl ester analogue (5b), with the highest
f value, has the lowest kcat within the 4-substituted
1-chloro-2-nitrobenzenes. The correlation with the charge
on C1 is directly a result of the electron-withdrawing
capacity of the 4-substituents: the higher this becomes,
the higher the positive charge on C1 becomes, and thus,
the higher the kcat. Regression analysis between the kcat
values of the 2-substituted 1-chloro-4-nitrobenzenes (1a -
10a ) and f gave an equation which was not significant (t
) -0.25), indicating that lipophilicity of the ortho sub-
stituents is not important for GSH conjugation by GST
4-4.
The Hammett σp constant, corrected for “through
resonance” (σp-), was also used in regression analyses
with the kinetic parameters ks and kcat/Km. For the GSH
conjugation of 2-substituted 1-chloro-4-nitrobenzenes
(1a -10a ), significant equations (t > |2|) were obtained
between ks and σp- and between kcat/Km and σp-, but with
low correlation coefficients (approximately 0.75). The
corresponding regression analyses for the 4-substituted
1-chloro-2-nitrobenzenes (1b-12b), however, gave equa-
tions with high correlation coefficients:
log ks ) 3.09((0.50)σp- - 5.43((0.46)
r ) 0.952 s ) 0.238 n ) 6
log kcat/Km ) 3.42((0.76)σp- - 4.13((0.70) (7)
r ) 0.913 s ) 0.365 n ) 6
(6)
Sp ecificity Con sta n ts kca t/Km a n d ks Va lu es. The
k
cat/Km parameter (specificity constant) for enzymes
comprises the processes of substrate binding, intermedi-
ate stabilization and release of products from the active
site. In addition to substrates 1b and 2b of the 4-sub-
stituted 1-chloro-2-nitrobenzenes, substrates with para
phenyl substituents (3b and 11b) also show substantial
GSH conjugation when catalyzed by base and by GST
4-4 (Table 2). Of both the 2-substituted 1-chloro-4-
nitrobenzenes (1a -10a ) and the 4-substituted 1-chloro-
2-nitrobenzenes (1b-12b), substrate 1a ,b (CDNB) has
the highest ks and kcat/Km values (Table 2) due to the high
electron-withdrawing capacity of the o- and p-NO2 sub-
stituents resulting in the best stabilization of the Meisen-
heimer intermediate in the present series of compounds.
In Figure 5 it is shown that all 4-substituted 1-chloro-
2-nitrobenzenes have higher ks values than the 2-sub-
stituted 1-chloro-4-nitrobenzenes, except for the tert-butyl
ester-substituted substrates 6a and 6b, which have
identical ks values. An overall comparison of kcat/Km
values of the GSH conjugation of 2-substituted 1-chloro-
As was observed for kcat, the process of “through reso-
nance” is obviously of more importance in the GSH
conjugation of 4-substituted 1-chloro-2-nitrobenzenes
than of 2-substituted 1-chloro-4-nitrobenzenes, because
of steric hindrance in the Meisenheimer intermediates
between the ortho substituent and the Cl atom at C1.
The slopes in the Hammett eqs 6 and 7 represent the
F values, which give an indication of the sensitivity of
the reaction for the electronic effects of the substituents.
Upon regression analysis between σp- and ks for the base-
catalyzed GSH conjugation of 4-substituted 1-chloro-2-
nitrobenzenes, Keen et al. found a F value of 3.1 when
using the substituents Cl, SO3-, CONH2, COCH3, CHO,
and NO2 (40), and Chen et al. found a F value of 3.4 (
0.2 when using the substituents SO3-, SO2CH3, COCH3,

446 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Van der Aar et al.
catalyzed reaction, suggests that there is a unique
interaction between the o-NO2 substituent and an amino
acid in the active site of GST 4-4. This interaction could
be of electronic nature, giving an extra stabilization
during formation of the Meisenheimer intermediates.
This observation confirms the previously proposed inter-
action between the 2-substituents of 2-substituted 1-chloro-
4-nitrobenzenes (IS2) and a Lewis acidic amino acid in
the active site of GST 4-4 (pIS2) (15).
E xt en sion of t h e Su b st r a t e Mod el for GST 4-4.
Descr ip tion of th e Su bstr a te Mod el for GST 4-4. On
the basis of the regio- and stereoselective GST 4-4-
catalyzed GSH conjugation reactions of 20 substrates
[aromatic diol epoxides, pyrene oxide, (aza)phenanthrene
oxides, and aromatic chlorides (i.a., substrates 1a -5a ,
9a , 10a , 4b, and 6b)], a substrate model has recently
been developed which comprises the main structural
characteristics of substrates for GST 4-4, using benzo-
[a]pyrene 7(R),8(S)-diol 9(S),10(R)-epoxide (BPDE) as a
template (Figure 1) (15). Several as yet hypothetical
structural elements in the active site of GST 4-4 have
been indicated: a flat aromatic region responsible for
orientation and stabilization of some substrates (pIS4)
and three other interaction sites consisting of Lewis acid
amino acids (pIS1, pIS2, and pIS3) (Figure 1). This
substrate model appeared to be useful to rationalize the
stereo- and regioselective GSH conjugation of a number
of substrates by fitting them in the model (styrene oxides,
stilbene oxides, 2-cyano-1,3-dimethyl-1-nitrosoguanidine,
and para-substituted 4-phenyl-3-buten-2-ones). More-
over, recently a protein homology model for rat GST 4-4
was developed (42), and in combination with the sub-
strate model for GST 4-4, various amino acids in or near
the active site could be indicated which might be respon-
sible for binding and/or activation of substrates.
F igu r e 6. Kinetic constants of the base-catalyzed (ks, b) and
the GST 4-4-catalyzed (kcat/Km, O) GSH conjugation of 4-sub-
stituted 1-chloro-2-nitrobenzenes vs σp-. Plots are based on eqs
6 and 7.
CHO, CF3, and NO2 (41). These values correspond well
to the F value found in the present study: 3.09 ( 0.50
(eq 6). For the GST 4-4-catalyzed GSH conjugation of
4-substituted 1-chloro-2-nitrobenzenes, Chen et al. found
a F value of 2.4 ( 0.6 (41), which does neither differ
significantly from the F value of 3.4 ( 0.8 found in the
present study (eq 7). When eqs 6 and 7 are transformed
to Hammett plots (Figure 6), it can also be seen that the
base-catalyzed as well as the GST 4-4-catalyzed reaction
with 4-substituted 1-chloro-2-nitrobenzenes depend on
-
σp in a similar way.
In a previous study, the ks and kcat/Km values of the
base- and GST 4-4-catalyzed GSH conjugation of 2-sub-
stituted 1-chloro-4-nitrobenzenes (1a -10a ) were cor-
related with σp (21). From the different F values in these
equations and additional regression analyses it was
concluded that the rate-limiting step in the GSH conju-
gation of 2-substituted 1-chloro-4-nitrobenzenes are dif-
ferent for the base-catalyzed and the GST 4-4-catalyzed
reactions (21). In the present study, however, no sig-
nificant difference in dependence on electronic properties
of the para substituents of 4-substituted 1-chloro-2-
nitrobenzenes is observed between the base- or GST 4-4-
catalyzed GSH conjugation reactions (Figure 6). Al-
though both types of substrates (1a -10a and 1b-12b)
are conjugated to GSH via an identical SNAr mechanism,
the nature and position of the substituents apparently
differentially influence the GSH conjugation rate and, by
inference, also the transition states.
Ad d ition of 2-Su bstitu ted 1-Ch lor o-4-n itr oben -
zen es to th e Su bstr a te Mod el. Like the o-methyl
ester- and o-butyl ester-substituted 1-chloro-4-nitroben-
zenes (4a and 5a , respectively) and the o-benzophenone
analogue 3a (15), substrate 6a could not be fitted on
BPDE in its minimal energy conformation, i.e., in which
the o-carbonyl oxygen atom lies approximately in the
plane of the aromatic ring. Therefore, these substrates
were fitted alternatively onto the BPDE template by
superimposing the sites of GS- attack and the carbonyl
oxygen atom of the substituents onto IS2. The energy
difference between the fitted and the minimal energy
conformation of substrate 6a was 5.53 kJ /mol, and the
distance between the fitted atoms was 0.05 Å, indicating
that this fit is acceptable. In this fit two of the methyl
groups of the tert-butyl ester substituent are positioned
on the border of an area where, in the substrate model
as yet, no atoms of superimposed substrates were found.
This observation, together with the fact that no measur-
able GSH conjugation of 6a was found up to 300 µM when
catalyzed by GST 4-4 while GSH conjugation of 6a was
observed when catalyzed by base (Table 2), suggests that
steric restrictions in this area of the active site might be
important. In a recent study, however, it was found that
6a competitively inhibited the GST 4-4-catalyzed GSH
conjugation of substrate 7a with a Ki of 32 µM, indicating
that 6a binds to the active site with a relatively high
affinity.2 This suggests that there are no steric restric-
tions in the binding of 6a to GST 4-4.
After multiple-regression analyses of log kcat/Km and
log ks of the 4-substituted 1-chloro-2-nitrobenzenes (1b-
12b) with all eight physicochemical and all five computer-
calculated parameters, only one significant correlation
was found, i.e., between log kcat/Km and the charge on the
-
o-NO2 substituent of the substrates and σp
:
log kcat/Km ) -24.47((4.41)(charge o-NO2) +
3.93((0.28)σp- - 8.63((0.84) (8)
r ) 0.993
s ) 0.125
n ) 6
-
The correlation coefficient with the single parameter σp
was significantly lower, 0.913, and the single regression
with the charge on o-NO2 showed no significant correla-
-
tion (t ) -0.18). Equation 8 implies that when the σp
increases and the charge on the o-NO2 substituent
becomes more negative, the kcat/Km will increase. The
fact that this correlation was only observed for the GST
4-4-catalyzed GSH conjugation and not for the base-
2
E. M. van der Aar et al., manuscript in preparation.

4-Substituted 1-Chloro-2-nitrobenzenes and Rat GST 4-4
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 447
Substrate 7a was fitted as well onto BPDE in its
minimal energy conformation by superimposing the sites
of GS- attack on one hand and the N-atom of the o-CN
substituent onto IS2 on the other hand. The distance
between the fitted atoms was 0.13 Å, which is indicative
for an acceptable fit. The o-CF3 analogue 8a was also
fitted onto the template molecule in its minimal energy
conformation. The sites of GS- attack were superim-
posed, and one of the F atoms of CF3 was fitted onto IS2,
leading to small distances between the fitted atoms (0.05
Å).
Ad d ition of 4-Su bstitu ted 1-Ch lor o-2-n itr oben -
zen es to th e Su bstr a te Mod el. For the fitting of
4-substituted 1-chloro-2-nitrobenzenes onto the template
molecule BPDE, the respective sites of attack by GSH
were matched and one of the oxygen atoms of the o-NO2
substituent were fitted onto IS2. The fits were performed
with the minimal energy conformations of the 4-substi-
tuted 1-chloro-2-nitrobenzenes, and the distance between
the fitted atoms was 0.11 Å. The methyl and tert-butyl
ester analogues 4b and 6b were previously used to probe
the steric restrictions of the active site of GST 4-4 (15).
Substrate 4b appeared to conjugate well to GSH, while
for 6b no measurable GSH conjugation was observed up
to 300 µM (Table 2). It was concluded that large bulky
substituents at the position para relative to C1 cannot
be accommodated in the active site of GST 4-4 (15).
The 4-halogen-substituted substrates 8b, 9b, and 10b
all fitted perfectly in the substrate model. No parts of
the para substituents were crossing areas, described by
the substrate model, and on the basis of their electron-
withdrawing capacity they should conjugate to GSH
when catalyzed by GST 4-4. Substrates 9b and 10b,
however, showed no measurable GSH conjugation, which
is probably due to the low extinction coefficients of the
corresponding GS-conjugates as discussed before and not
to steric restrictions (Br and CF3 have similar sizes, rvdw
) (1.85 Å).
Substrates 2b and 5b, similar to 4b as previously
described, can be fitted in two conformations (see Figure
2A and 2B) and apparently, because of their substantial
GSH conjugation rate (Table 2), can be accommodated
in either of the two conformations. For substrate 5b
several conformations are possible, because of the long
flexible butyl chain. The lowest sterical hindrance, and
by inference the lowest total energies, is observed in the
extended conformations. Of the two extended conforma-
tions, the lowest energy conformations were calculated
and used for a more detailed description of the substrate
model for GST 4-4. Obviously, there is quite some space
in the area para relative to the GSH conjugation site of
4-substituted 1-chloro-2-nitrobenzenes, because the long
butyl ester part is not negatively influencing the GSH
conjugation by GST 4-4. It cannot be excluded, however,
that the GST protein accommodates the n-butyl chain
in a folded conformation.
From the benzophenone- and phenyl ester-substituted
compounds (3b and 11b, respectively), which both have
a high apparent affinity for GST 4-4 and a relatively high
GSH conjugation rate (Table 2), four low-energy confor-
mations were considered. Both substrates have two
orientations of the carbonyl (Figure 2) and because of
intramolecular sterical hindrance, the phenyl rings are
positioned approximately +65° and -65° out of the
aromatic plane for 3b, while this is approximately +58°
and -58° for 11b. In Figure 7 are shown the fits of the
four possible low energy conformations of 3b and 11b
onto BPDE. In the area para relative to the site of GSH
attack (C1), a shaft becomes visible between the two
mirrored conformations. In addition to extended alkyl
chains at the para position, also relatively flat para
substituents are obviously accommodated by GST 4-4.
To further examine the steric restrictions of para
substituents for GST 4-4 and the importance of the above-
mentioned shaft, the 3′-tert-butylphenyl ester-substituted
compound 12b was synthesized and tested as a substrate
for GST 4-4. Up to a concentration of 300 µM, no
formation of GS-conjugate was measured. Electronic
restrictions play no role because this compound was
conjugated to GSH when catalyzed by base (Table 2).
Eight low-energy conformations were fitted: the two
orientations of the carbonyl (Figure 2), the phenyl ring
twisted above and under the aromatic plane due to steric
hindrance, and the tert-butyl substituent at both sides
of the phenyl ring. In Figure 7 these eight conformations
are fitted onto BPDE. It is clear that the tert-butyl group
occupies the above mentioned shaft, and this might imply
steric restrictions in this area of the active site of GST
4-4. To confirm this theory, the two low-energy confor-
mations of the tert-butyl ester analogue 6b, which was
not conjugated to GSH up to 300 µM by GST 4-4, were
again fitted on the template (Figure 7). Also in this case,
the bulky tert-butyl group occupies the defined shaft, and
this might therefore also be a reason for 6b not being a
substrate for GST 4-4. These observations are in line
with the previously discussed SAR equation 4. When the
B5 Sterimol parameter becomes too large, as in the case
of tert-butyl ester and tert-butyl phenyl ester substitu-
ents, the kcat values become so low that they are no longer
measurable.
We are aware of the fact that there might in principle
be alternative binding modes of the 4-substituted 1-chloro-
2-nitrobenzenes to the active site of GST 4-4. Several of
these alternative binding modes have been considered,
but the binding modes presented appeared to confirm the
obtained results in the best possible way.
Con clu sion s
In the present study, quantitative structure-activity
relationships (SAR’s) are described for base- and GST 4-4-
catalyzed GSH conjugation of 4-substituted 1-chloro-2-
nitrobenzenes (1b-12b). Eight physicochemical param-
eters, delineating electronic, steric, and lipophilic effects
of the substituents, as well as five computer-calculated
molecular parameters of the substrates (charge distribu-
tions and energy values) were used in the regression
analyses with the kinetic parameters for the base- (ks)
and GST 4-4-catalyzed (Km, kcat, kcat/Km) reactions. The
obtained SAR’s were compared with corresponding SAR’s
for 2-substituted 1-chloro-4-nitrobenzenes (1a-10a). From
a previous study, it appeared that the ks and kcat/Km
values of the base- and GST 4-4-catalyzed GSH conjuga-
tion of 2-substituted 1-chloro-4-nitrobenzenes depend
differentially on the electronic properties of the 2-sub-
stituents. This suggested that these substrates are
conjugated to GSH via different rate-limiting transition
states when catalyzed by base or GST 4-4 (21). In the
present study, however, the base- and GST 4-4-catalyzed
GSH conjugation of 4-substituted 1-chloro-2-nitroben-
zenes showed no significant differences in the dependence
on the electronic properties of the substituents. This
indicates that the position of the substituents relative
to the reaction center and their electron-withdrawing

448 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Van der Aar et al.
F igu r e 7. Fits of 4-substituted 1-chloro-2-nitrobenzenes onto the template molecule BPDE: three conformations of 3b (upper left),
four conformations of 11b (upper right), eight conformations of 12b (lower left), and two conformations of 6b (lower right). Hydrogen
atoms have been omitted.
capacity determine whether the rate-limiting transition
state for the base-catalyzed nucleophilic aromatic sub-
stitution reaction is the same as for the GST 4-4-
catalyzed reaction. Log ks, log kcat, and log kcat/Km of the
GSH conjugation of 4-substituted 1-chloro-2-nitroben-
zenes (1b-12b) correlated well with the Hammett σp
constant, corrected for through resonance (σp-), while the
corresponding parameters of the 2-substituted 1-chloro-
4-nitrobenzenes (1a-10a) did not. The process of through
resonance is obviously of more importance in the GSH
conjugation of 4-substituted 1-chloro-2-nitrobenzenes
than of 2-substituted 1-chloro-4-nitrobenzenes, because
in the latter series of substrates this process is probably
not efficient, due to steric hindrance in the Meisenheimer
intermediates between the ortho substituent and the Cl
atom at C1. Log kcat of GST 4-4 for the GSH conjugation
of 2-substituted 1-chloro-4-nitrobenzenes correlated highly
with the charge on the p-NO2 substituent of these
substrates, which appeared to be an indirect parameter
for the electron-withdrawing capacity of the 2-substitu-
ents, corrected for steric influences (σo).
described by De Groot et al. (15). Addition of these
substrates gave more detailed information about the
steric restrictions of substrates of GST 4-4 and by
inference about the active site of GST 4-4. For the
substrates with relative bulky substituents, tert-butyl
ester and 3′-tert-butylphenyl ester, GSH conjugation was
observed when catalyzed by base but not by GST 4-4.
These compounds were fitted in the substrate model and
appeared to occupy an area in the substrate model which
can be assigned as a “forbidden area”.
Refer en ces
(1) Commandeur, J . N. M., Stijntjes, G. J ., and Vermeulen, N. P. E.
(1995) Enzymes and transport systems involved in the formation
and disposition of glutathione S-conjugates. Pharmacol. Rev. 47,
271-330.
(2) Armstrong, R. N. (1991) Glutathione S-transferases: Reaction
mechanism, structure, and function. Chem. Res. Toxicol. 4, 131-
140.
(3) Mannervik, B., A° lin, P., Guthenberg, C., J enssen, H., Tahir, M.
K., Warholm, M., and J ornvall, H. (1985) Identification of three
classes of glutathione transferases common to several mammalian
species. Correlation between structural data and enzymatic
properties. Proc. Natl. Acad. Sci. U.S.A. 82, 7202-7206.
Multiple-regression analyses with kcat and kcat/Km for
the GSH conjugation of 4-substituted 1-chloro-2-nitroben-
zenes showed that, besides electronic effects, also steric
(L and B5) and lipophilic (f) factors of the substituents
play a role. The lipophilicity of the para substituents,
for example, should not be too high, because if it is, the
GSH conjugation rate will be low.
(4) Meyer, D. J ., Coles, B., Pemble, S. E., Gilmore, K. S., Fraser, G.
M., and Ketterer, B. (1991) Theta, a new class of glutathione
transferases purified from rat and man liver. Biochem. J . 274,
409-414.
(5) Buetler, T. M., and Eaton, D. L. (1992) Glutathione S-trans-
ferases: Amino acid sequence comparison, classification and
phylogenetic relationship. Environ. Carcinog. Ecotoxicol. Rev.
C10, 181-203.
The 4-substituted 1-chloro-2-nitrobenzenes were also
used to refine the substrate model for GST 4-4, recently

4-Substituted 1-Chloro-2-nitrobenzenes and Rat GST 4-4
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 449
(6) J ohnson, W. W., Liu, S. X., J i, X. H., Gilliland, G. L., and
Armstrong, R. N. (1993) Tyrosine-115 participates both in chemi-
cal and physical steps of the catalytic mechanism of a glutathione
S-transferase. J . Biol. Chem. 268, 11508-11511.
(7) Ploemen, J . H. T. M., J ohnson, W. W., J espersen, S., Vanderwall,
D., Van Ommen, B., Van der Greef, J ., Van Bladeren, P. J ., and
Armstrong, R. N. (1994) Active-site tyrosyl residues are targets
in the irreversible inhibition of a class mu glutathione transferase
by 2-(S-glutathionyl)-3,5,6-trichloro-1,4-benzoquinone. J . Biol.
Chem. 269, 26890-26897.
(8) Zhang, P., Liu, S., Shan, S., J i, X., Gilliland, G. L., and Armstrong,
R. N. (1992) Modular mutagenesis of exons 1, 2, and 8 of a
glutathione S-transferase from the Mu class. Mechanistic and
structural consequences for chimeras of isoenzyme 3-3. Biochem-
istry 31, 10185-10193.
(9) Katusz, R. M., Bono, B., and Colman, R. F. (1992) Identification
of Tyr115 labeled by S-(4-bromo-2,3-dioxobutyl)glutathione in the
hydrophobic substrate binding site of glutathione S-transferase,
isoenzyme 3-3. Arch. Biochem. Biophys. 298, 667-677.
(10) Cooke, R. J ., Bjo¨rnestedt, R., Douglas, K. T., McKie, J . H., King,
M. D., Coles, B., Ketterer, B., and Mannervik, B. (1994) Photo-
affinity labelling of the active site of the rat glutathione trans-
ferases 3-3 and 1-1 and human glutathione transferase A1-1.
Biochem. J . 302, 383-390.
(11) Penington, C. J ., and Rule, G. S. (1992) Mapping the substrate-
binding site of a human class mu glutathione transferase using
nuclear magnetic resonance spectroscopy. Biochemistry 31, 2912-
2920.
(12) Reinemer, P., Dirr, H. W., Ladenstein, R., Scha¨ffer, J ., Gallay,
O., and Huber, R. (1991) The three-dimensional structure of class
π glutathione S-transferase in complex with glutathione sulfonate
at 2.3 Å resolution. EMBO J . 10, 1997-2005.
(13) J i, X., Zhang, P., Armstrong, R. N., and Gilliland, G. L. (1992)
The three-dimensional structure of glutathione S-transferase from
the Mu gene class. Structural analysis of the binary complex of
isoenzyme 3-3 and glutathione at 2.2-Å resolution. Biochemistry
31, 10169-10184.
(14) Sinning, I., Kleywegt, G. J ., Cowan, S. W., Reinemer, P., Dirr, H.
W., Huber, R., Gilliland, G. L., Armstrong, R. N., J i, X. H., Board,
P. G., Olin, B., Mannervik, B., and J ones, T. A. (1993) Structure
determination and refinement of human-alpha class glutathione
transferase A1-1, and a comparison with the Mu-class and Pi-
class enzymes. J . Mol. Biol. 232, 192-212.
(15) De Groot, M. J ., Van der Aar, E. M., Nieuwenhuizen, P. J ., Van
der Plas, R. M., Donne´-Op den Kelder, G. M., Commandeur, J .
N. M., and Vermeulen, N. P. E. (1995) A predictive substrate
model for rat glutathione S-transferase 4-4. Chem. Res. Toxicol.
8, 649-658.
(16) Adang, A. E. P., Moree, W. J ., Brussee, J ., and Mulder, G. J . (1991)
Inhibition of glutathione S-transferase 3-3 by glutathione deriva-
tives that bind covalently to the active site. Biochem. J . 278, 63-
68.
(17) Danielson, U. H., Esterbauer, H. and Mannervik, B. (1987)
Structure-activity relationships of 4-hydroxyalkenals in the
conjugation catalysed by mammalian glutathione transferases.
Biochem. J . 247, 707-713.
(18) Rietjens, I. M. C. M., Soffers, A. E. M. F., Hooiveld, G. J . E. J .,
Veeger, C., and Vervoort, J . (1995) Quantitative structure-
activity relationships based on computer calculated parameters
for the overall rate of glutathione S-transferase catalyzed conju-
gation of a series of fluoronitrobenzenes. Chem. Res. Toxicol. 8,
481-488.
(19) Soffers, A. E. M. F., Ploemen, J . H. T. M., Moonen, M. J . H.,
Wobbes, T., Van Ommen, B., Vervoort, J ., Van Bladeren, P. J .,
and Rietjens, I. M. C. M. (1996) Regioselectivity and quantitative
structure-activity relationships for the glutathione conjugation
of a series of fluoronitrobenzenes by purified glutathione S-
transferase enzymes from rat and man. Chem. Res. Toxicol. 9,
638-646.
(20) Van der Aar, E. M., Buikema, D., Commandeur, J . N. M., Te
Koppele, J . M., Van Ommen, B., Van Bladeren, P. J ., and
Vermeulen, N. P. E. (1996) Enzyme kinetics and substrate
selectivities of rat glutathione S-transferase isoenzymes towards
a series of new 2-substituted 1-chloro-4-nitrobenzenes. Xenobiotica
26, 143-155.
(21) Van der Aar, E. M., De Groot, M. J ., Bijloo, G. J ., Van der Goot,
H., and Vermeulen, N. P. E. (1996) Structure-activity relation-
ships for the glutathione conjugation of 2-substituted 1-chloro-
4-nitrobenzenes by rat glutathione S-transferase 4-4. Chem. Res.
Toxicol. 9, 527-534.
(22) Hammond, G. S., and Modic, F. J . (1953) Aromatic nitration. I.
The ultraviolet spectra of aromatic nitro compounds in sulfuric
acid. J . Am. Chem. Soc. 75, 1385-1388.
(23) Chemical Design Ltd. (1993) ChemX, Version J anuary 1993.
(24) Department of Chemistry Columbia University (1994) Macro-
model, Version 4.5.
(25) Weiner, S. J ., Kollman, P. A., Case, D. A., Singh, U. C., Ghio, C.,
Alagona, G., Profeta, S., and Weiner, P. (1984) A new force field
for molecular mechanical simulation of nucleic acids and proteins.
J . Am. Chem. Soc. 106, 765-784.
(26) Weiner, S. J ., Kollman, P. A., Nguyen, D. T., and Case, D. A.
(1986) An all-atom force field for simulations of proteins and
nucleic acids. J . Comput. Chem. 7, 230-252.
(27) Department of Chemistry Columbia University (1994) Batchmin,
Version 4.0.
(28) Dupuis, M., Spangler, D., and Wendoloski, J . (1980) NRCC
Program No. QG01 (GAMESS).
(29) Guest, M. F., van Lenthe, J . H., Kendrick, J ., Schoffel, K.,
Sherwood, P., Harrison, R. J ., with contributions from: Amos, R.
D., Buenker, R. J ., Dupuis, M., Handy, N. C., Hillier, I. H.,
Knowles, P. J ., Bonacic-Koutecky, V., von Niessen, W., Saunders,
V. R., and Stone, A. J . (1993) GAMESS-UK, IBM RS6000 Version
2.1.
(30) Hehre, W. J ., Stewart, R. F., and Pople, J . A. (1969) Self-consistent
molecular orbital methods. I. Use of Gaussian expansions of
Slater-type atomic orbitals. J . Chem. Phys. 51, 2657-2664.
(31) Stone, A. J . (1981) Distributed multipole analysis, or how to
describe a molecular charge distribution. Chem. Phys. Lett. 83,
233-239.
(32) Binkley, J . S., Pople, J . A., and Hehre, W. J . (1980) Self consistent
molecular orbital methods. 21. Small split-valence basis sets for
first-row elements. J . Am. Chem. Soc. 102, 939-947.
(33) Gordon, M. S., Binkley, J . S., Pople, J . A., Pietro, W. J ., and Hehre,
W. J . (1982) Self consistent molecular orbital methods. 22. Small
split-valence basis sets for second-row elements. J . Am. Chem.
Soc. 104, 2797-2803.
(34) Silipo, C., and Vittoria, A. (1990) Three-dimensional structure of
drugs. In Comprehensive medicinal chemistry (Hansch, C., Sammes,
P. G., and Taylor, J . B., Eds.) Vol. 4, pp 153-204, Pergamon Press,
Oxford.
(35) Rekker, R. F., and Mannhold, R. (1992) Calculation of drug
lipophilicity. The hydrophobic fragmental constant approach,
VCH, Weinheim, Germany.
(36) Hansch, C., and Leo, A. (1979) Substituent constants for correla-
tion analysis in chemistry and biology, Wiley, New York.
(37) Hammett, L. P. (1940) Physical Organic Chemistry, McGraw-Hill,
New York.
(38) Swain, C. G., and Lupton, E. C. (1968) Field and resonance
components of substituent effects. J . Am. Chem. Soc. 90, 4328-
4337.
(39) Miller, J . (1968) in Reaction mechanisms in organic chemistry
(Eaborn, C., and Chapman, N. B., Eds.) Vol. 8, pp 137-179,
Elsevier, New York.
(40) Keen, J . H., Habig, W. H., and J akoby, W. B. (1976) Mechanism
for the several activities of the glutathione S-transferases. J . Biol.
Chem. 251, 6183-6188.
(41) Chen, W.-J ., Graminski, G. F., and Armstrong, R. N. (1988)
Dissection of the catalytic mechanism of isoenzyme 4-4 of
glutathione S-transferase with alternative substrates. Biochem-
istry 27, 647-654.
(42) De Groot, M. J ., Vermeulen, N. P. E., Mullenders, D. L. J ., and
Donne´-Op den Kelder, G. M. (1996) A homology model for rat Mu
class glutathione S-transferase 4-4. Chem. Res. Toxicol. 9, 28-
40.
(43) Habig, W. H., and J akoby, W. B. (1981) Assay for differentiation
of glutathione S-transferases. Methods Enzymol. 77, 398-405.
TX960137W